From: WarmSpgs@aol.com
Message-ID: <28.10bf714c.27af5ff2@aol.com>
Date: Sun, 4 Feb 2001 20:46:26 EST
Subject: Re: LF: PLL inaccuracy
To: rsgb_lf_group@blacksheep.org
Cc: akestelo@bellatlantic.net
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org
Sender: <majordom@post.thorcom.com>

In a message dated 2/4/01 2:16:24 PM Eastern Standard Time, 
akestelo@bellatlantic.net writes:

<< Allow me to quote from  the book "Phase Locked Loops, Principle and  
Practice",  McGraw Hill 1996, by P.V. Brennan University College, London.  
Page 22: "...As far as the control loop is concerned, it should be noted that 
frequency dividers act equally as phase dividers, so that a factor of  1/N 
must be allowed for in the loop equations". >>

This is certainly true.  It must be taken into account in designing a loop, 
but it does not represent a source of _frequency_ error so long as lock is 
maintained.

The example cited from page 128 is such an extreme case that, were the 
controlled oscillator sufficiently unstable or noisy as to experience 
short-term excursions beyond 360 degrees at its own frequency--or if the 
detector were occasionally unable to resolve with the 0.35 degree precision 
specified--the entire system would be susceptible to coming out of lock 
entirely!  Until that point, the output is still on-frequency, but might 
exhibit serious phase noise.

This is why one must perform phase comparisons at a reasonable division 
ratio, even if it means sacrificing extremely narrow frequency steps; or 
else, one has to be much more meticulous in the design of the VCO and the 
phase comparator, and even then, lockup times may be rather long.

73,
John